Three experiments were conducted to compare the excretion of water-soluble phosphorus (P) of starter, grower, and finisher pigs fed corn-soybean meal-based P-deficient basal diet containing no added inorganic P (B), P-adequate diet (the basal diet with added dicalcium phosphate; B + P), the basal diet plus 500 units of microbial phytase/kg (B + 500), or the basal diet plus 1000 units of microbial phytase/kg (B + 1000). There were 6 barrows per diet, in a randomized complete block design, with an average initial body weight of 10.4, 20.0, or 51.1 kg for each of starter, grower, and finisher pigs in the three phosphorus balance experiments, respectively. In the experiments, the addition of dicalcium phosphate or phytase to the basal diet increased (P < 0.05) the digestibility and retention of P, and there were linear reductions (P < 0.05) in water-soluble P excretion as a result of supplementing the basal diet with phytase. In the starter pig experiment, phytase addition at 500 or 1000 units/kg reduced (P < 0.05) water-soluble P excretion by 28 or 42%, respectively when compared with the B + P diet. In the grower pig experiment, adding phytase at 500 or 1000 units/kg reduced (P < 0.05) water-soluble P excretion by 24 or 34%, respectively when compared with the B + P diet. The use of phytase at 500 or 1000 units/kg reduced (P < 0.05) water-soluble P excretion by 11 or 30%, respectively in the finisher pig experiment. The proportion of water-soluble phosphorus in total phosphorus was not affected by dietary treatment in any of the three experiments. In conclusion, adding phytase at 1000 units/kg to a corn-soybean meal-based P-deficient diet basal diet containing no added inorganic P compared with B + P diet reduced the daily excretion of water-soluble P in starter, grower, and finisher pigs by 42, 34, and 30%, respectively.